CHAPTER III. SYNTHESIS OF THE BENZYLISOQUINOLINE ALKALOID
3.2. Results
3.2.2. Production of dopamine
Two different pathways for the production of dopamine that incorporate the TYDC2 activity are conceivable. One pathway requires a phenyloxidase activity to convert tyrosine to L-dopa while the other requires a phenolase activity to convert tyramine to dopamine (Fig. 3.3).
HO
OH NH2
O
HO NH2
HO HO
NH2 L-Tyrosine TYDC
L-dopa
Dopamine HO
OH NH2
O HO
Tyramine 1 TYDC
2
Fig. 3.3. Alternative pathways for dopamine production from tyrosine. The upper pathway (shown in red) uses an unnamed enzyme activity (1) to convert tyrosine to L- dopa and the TYDC activity to convert L-dopa to dopamine. The lower pathway (shown in blue) uses the TYDC activity to convert tyrosine to tyramine and a second enzyme activity (2) to convert tyramine to dopamine.
The oxidation of tyrosine to L-dopa can be catalyzed by three different types of enzymes: tyrosine hydroxylase, tyrosinase, and β-tyrosinase43. Tyrosinase activities are widely distributed in nature as they are key enzymes involved in melanin biosynthesis.
However, in addition to catalyzing the orthohydroxylation of phenolic substrates, they also catalyze the oxidation of these catechol products to quinones. In the host organism and in the presence of activating nucleophiles, quinones are spontaneously polymerized to form melanin. Reactive quinone intermediates have antibiotic properties and melanin
itself exerts many strengthening and protective effects; however, accumulation of the L- dopa intermediate is desired in this case, contrary to what happens in nature.
Yeast strains were constructed to express either the human tyrosine hydroxylase 2 (hTH2) or rat tyrosine hydroxylase (TyrH) along with the human GTP cyclohydrolase I (hGTPCHI) required to synthesize the cofactor tetrahydrobiopterin (BH4). Preliminary results indicated no L-dopa accumulation in the growth media or cell extracts of these strains. In addition, expression levels of these recombinant proteins were assayed by Western blotting using a standard epitope tag, and all three fell below the detection threshold. Difficulties with expression coupled to the fact that hTH2 in particular has a tyrosine hydroxylase to dopa oxidase ratio of ~2:144 led us to pursue other enzymes for this activity. However, methods such as codon-optimization, use of more favorable TH variants (such as hTH4), and optimization of the cofactors Cu(II) and BH4 may show more encouraging results.
We also obtained sequences coding for two tyrosinase cDNAs from Agaricus bisporus (AbPPO1 and AbPPO2), the common button mushroom. Results from AbPPO2 expressed in yeast were encouraging but inconsistent. One strain was able to show production of dopamine when co-transformed with a plasmid expressing TYDC2.
However, the addition of 1-10 μM Cu(II)SO4, which should facilitate the catalytic activity of the copper-containing tyrosinase enzyme, increased accumulation of L-dopa but produced no dopamine. TYDC2 remained active in the presence of Cu(II) as evidenced by tyramine accumulation but failed to convert L-dopa to dopamine under these conditions (Fig. 3.4).
0 1 2 3 4 5 6 7 8 0
1 2 3 4 5
0 1 2 3 4 5 6 7 8
0 1 2 3
Time (min)
Intensity (ion counts)
Dopamine
L-dopa
Tyramine
Tyramine a
b
Fig. 3.4. LC-MS analysis of the growth media of CSY88 (Table 3.1) expressing AbPPO2 and TYDC2. (a) Production of tyramine (m/z = 138, blue) and dopamine (m/z = 154, green) are observed in synthetic complete media with no additives. (b) Production of tyramine (m/z = 138, blue) and L-dopa (m/z = 198) are observed when 5 μM Cu(II)SO4 is added to the media; dopamine production is not detectable.
In an effort to increase metabolite shuttling between AbPPO2 and TYDC2 by co- localizing the enzymes in vivo, a fusion protein between the two enzymes was constructed and the proteins were tagged with matching C-terminal or N-terminal leucine zippers. In these strains, tyramine production remained high while L-dopa production was nearly abolished, and no dopamine was produced. As proteolytic processing of AbPPO2 in Agaricus is believed to take place from the C-terminal region45, additional amino acids may interfere with processing and maturation of the protein. This is true specifically in the case of the fusion protein as it was constructed with the AbPPO2 domain on the N-terminus. Also, the C-terminal leucine zipper sequence would likely be cleaved during post-translational processing. Evidence for extensive proteolytic processing is provided by Western blotting analysis using N- and C-terminal tags (Fig.
3.5). However, the number of discrete bands suggests that the desired product does not dominate in our host. Differences in signal sequences and proteases between the two
AbPPO2 (N)
AbPPO2 (C)
Fig. 3.5. Western blots of AbPPO2 constructs. Left: AbPPO2 with an N-terminal tag detected with an Anti-His G-HRP antibody. Right: AbPPO2 with a C-terminal tag detected with an Anti-V5 HRP antibody.
An alternative pathway for dopamine production proceeds through tyramine and requires the enzymatic addition of a 3’-hydroxyl group. The most well-characterized enzyme that performs this reaction is the human cytochrome P450 2D6 (hCYP2D6).
While yeast is a suitable and often-used host for P450 expression, this class of enzymes still presents many difficulties as will be demonstrated throughout this work.
Overexpression of the endogenous yeast NADPH cytochrome P450 reductase (yCPR1) or other heterologous P450 reductase is generally required to observe activity of P450 enzymes either in vivo or using yeast microsomal preparations. Among the most commonly used strains for expression of heterologous P450s are W(R) and WAT11 strains which contain genomic integrations of GAL-inducible yeast CPR1 and Arabidopsis thaliana ATR1 reductases, respectively46. We tested both strain backgrounds with co-expression of the TYDC2 and hCYP2D6 enzymes for dopamine production. The engineered yeast cells in the WAT11 strain background showed no dopamine production when grown in the presence of galactose, indicating that the A. thaliana ATR1 is not a
suitable reductase partner for hCYP2D6. The W(R) strain background produced dopamine, but surprisingly, no difference was observed between induced and uninduced cultures. This suggests that leaky expression of yCPR1 from the GAL1-10 promoter is sufficient to enhance the activity of its P450 partner above that seen in the wild-type background strain. Additional expression of the yCPR1 driven by the tetO2 promoter (which is constitutive in the absence of doxycycline) further increased dopamine production. The use of rich media (2x yeast nitrogen base and 5 g l-1 casein hydrolysate instead of ammonium sulfate) also boosted dopamine production slightly but is not an economical solution for an industrial process.
We sought to improve CYP2D6 expression by optimizing codon usage of the gene for yeast. We were not able to observe either improved expression or activity but retained this sequence to reduce metabolic stress on the cells. We co-expressed this optimized sequence with the soluble yCPRΔ33 and also constructed a fusion protein between yCYP2D6 and yCPRΔ33. This fusion construct yielded the highest dopamine levels when expressed from a strong TEF6 promoter in a strain background with either an integrated or plasmid-based copy of TYDC2. This engineered yeast strain produced ~10 μM dopamine while other strains fell short of this mark (Fig. 3.6). Although experimental
evidence suggests that this level should be sufficient for incorporation into norcoclaurine, we have not been able to demonstrate this in vivo by supplementing 4-HPA. Aside from the redox balance, many other factors may be limiting, including the low affinity of CYP2D6 for the substrate tyramine and autooxidation of dopamine. Additional protein evolution, strain engineering, and optimization of media conditions are required to address these potential issues.
1 . 5 2 .0 2 . 5 3 . 0 3 .5 4 . 0 4 . 5 5 .0 5 . 5 0 . 0
0 . 5 1 . 0 1 . 5 2 . 0 2 . 5
1 . 5 2 .0 2 . 5 3 . 0 3 .5 4 . 0 4 . 5 5 .0 5 . 5
0 . 0 0 . 5 1 . 0 1 . 5 2 . 0 2 . 5
Time (min)
Intensity (ion counts)
x105
Fig. 3.6. LC-MS analysis of dopamine produced in yeast. Production of dopamine by engineered yeast strains expressing PTEF1:TYDC2 and PTEF6:yCYP2D6-yCPRΔ33 from high-copy plasmids (red; pCS221 and pCS1565) compared to a 10 μM dopamine standard (black). Extracted ion chromatograms for m/z = 154 are shown.
3.2.3. Production of 4-hydroxyphenylacetaldehyde and 3,4-dihydroxyphenylacetaldehyde